blob: 33a96e9c338ce9ee66d06082693947e42bae98b0 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
\section{Background}
Compiler Fuzzing is a popular technique to find bugs in programs, and it is especially effective at finding compiler bugs, as Yang \textit{et al.}~\cite{yang11_findin_under_bugs_c_compil} demonstrated by finding more than 300 bugs in GCC and clang. This technique can therefore also be used to find bugs in HLS tools. There has also been some prior work in trying to find bugs in HLS tools and also ensuring that synthesis tools output a correct design.
Lidbury et al. \cite{lidbury15_many_core_compil_fuzzin} fuzz-tested several OpenCL compilers, including an HLS compiler from Altera (now Intel). They were only able to subject that compiler to superficial testing because so many of the test cases they generated led to it crashing. In comparison to our work: where Lidbury et al. generated target-independent OpenCL programs that could be used to test HLS tools and conventional compilers alike, we specifically generate programs that are tailored for HLS (e.g. with HLS-specific pragmas) with the aim of testing the HLS tools more deeply. Another difference is that where we test using sequential C programs, they test using highly concurrent OpenCL programs, and thus have to go to great lengths to ensure that any discrepancies observed between compilers cannot be attributed to the nondeterminism of concurrency.
Herklotz et al.~\cite{verismith} fuzz-tested several FPGA synthesis tools using randomly generated Verilog programs. Where they concentrated on the RTL-to-netlist stage of hardware design, we focus our attention on the earlier C-to-RTL stage.
Several authors have taken steps toward more rigorously engineered HLS tools that may fare better under testing campaigns like ours. One such example might be a Handel-C compiler which has been partially formally proven in a theorem prover called HOL~\cite{perna12_mechan_wire_wise_verif_handel_c_synth}. However, Handel-C requires the manual parallelisation of hardware and would require many more checks to avoid data races. In addition to that, a HLS tool called SPARK~\cite{gupta03_spark} also contains passes that are validated during compilation time~\cite{chouksey20_verif_sched_condit_behav_high_level_synth}, such as the scheduling algorithm. Unfortunately the tools are not readily available yet to properly test, or do not support C as input directly. Finally Catapult C~\cite{mentor20_catap_high_level_synth} is also a high-level synthesis tool that proves that the netlist is equivalent to the input, and should therefore not produce any bugs and only fail if the output cannot be proven.
%%% Local Variables:
%%% mode: latex
%%% TeX-master: "main"
%%% TeX-command-extra-options: "-shell-escape"
%%% End:
|
